This music sounds like the visual equivalent of one of those motion settings on a tv that cause the "Soap Opera Effect". I'm wondering if the LLM is failing to emulate the odd and even harmonics of real recordings and that's why it seems so unnatural. It makes me feel sick. It's different than the sterile straight to digital recordings of the late 80s early 90s too. Do any of these let people download the individual…
Suno lets you download the stems, also no affiliation at all just really like the tool and have spent a lot of time with it: Suno is night and day better than this product, by a mile. https://soundcloud.com/charlielabs/charlie (everything on this account I made with Suno, sometimes through Abelton if I'm not feeling lazy)
It's interesting to me people really hate AI gen'd music, some people I know literally won't listen to it and just like... get kinda angry I even tried to show them what I made (so I stopped showing people for now). I have a good memory for sound, but whatever neurodiversity I have makes learning instruments (and math, and written words, and and) super challenging. I loved screwing around remixing stuff in Abelton, b…
> what I made
But did you really make it? Or was it the AI? If someone commissions a piece of art from an artist, I don't think the commissioner would be able to truthfully say they made it, even if they had a specific vision for what the piece of art should look like. But if you've edited or changed the track enough yourself, maybe it would be fair to call yourself a co-author...
I've played guitar for 23 years, and there is something just off-putting about most of the music on that page, but particularly "Yellow Bus Jam". The guitar solo sounds very unnatural, especially the phrasing, which is totally random. Blues musicians are actually attempting to say something through their instrument. This was just a random number generated solo played by a 6 finger three handed robot. No thanks, lol.
Get used to it, because it's much cheaper than a musician, and to the average person "attempting to say something through their instrument" and "random number generated solo" are largely the same thing.
As a non-music person, can confirm - if someone tried to tell me something through their instrument I would probably tell them it should have been an email.
I can generally understand that music has moods, but don’t think I could distinguish human-generated music from silicon-generated music at this point (unless I recognize a specific artist, of which there are vanishingly few I’m capable of)
I could be wrong, but I think the use case here is mainly for non-artists in domains where the music is not particularly important. For example, a podcaster/youtuber may want a short intro track. An entertainer or a marketer may want some generic or silly background music. Does it have a use case for a producer/musician? Maybe. It might give them ideas for chord progressions, melodies, etc. But real music does that t…
Polemics about generative AI might indeed benefit from separately addressing art and entertainment. Generative AI in art is worth debating, but entertainment is not even a question: entertainment is a proven market for mass-produced slop, where artless works just fine and art is marginally valued.
Get used to it, because it's much cheaper than a musician, and to the average person "attempting to say something through their instrument" and "random number generated solo" are largely the same thing.
As a non-music person, can confirm - if someone tried to tell me something through their instrument I would probably tell them it should have been an email. I can generally understand that music has moods, but don’t think I could distinguish human-generated music from silicon-generated music at this point (unless I recognize a specific artist, of which there are vanishingly few I’m capable of)
It's interesting to me people really hate AI gen'd music, some people I know literally won't listen to it and just like... get kinda angry I even tried to show them what I made (so I stopped showing people for now). I have a good memory for sound, but whatever neurodiversity I have makes learning instruments (and math, and written words, and and) super challenging. I loved screwing around remixing stuff in Abelton, b…
> what I made But did you really make it? Or was it the AI? If someone commissions a piece of art from an artist, I don't think the commissioner would be able to truthfully say they made it, even if they had a specific vision for what the piece of art should look like. But if you've edited or changed the track enough yourself, maybe it would be fair to call yourself a co-author...
I'm not going to defend against that position generally. Your point is well taken, I don't know if I put enough into it to consider it real to you. I think about the lyrics, I write the MVP of them by hand, and then I have 4.5 give me variations of them, I write all my instrumental prompts by hand I don't use any LLM for that because I typically know exactly what I want and I tend to be quite precise with my instructions there. Suno allows you a lot of fine grain control, and you can also create personas from tracks and then use those personas in new tracks so as you can imagine you can spend a long time trying to get exactly what you want... last night I gen'd over 60 songs before I got it dialed in how I wanted it and I'm still not totally happy with it. I understand the position though, I was on original team that took deviantart from zero to millions of artists through the early 2000s and have a few Emmy awards from the film to digital transition days, I'm no stranger to these conversations.
Get used to it, because it's much cheaper than a musician, and to the average person "attempting to say something through their instrument" and "random number generated solo" are largely the same thing.
I'm not anti-AI, but I strongly believe the human element of music can be imitated but not fully replicated. Listening back to that song I can hear the attempt to stylistically play slightly off-beat to get the feel of a band playing without a metronome. The auditory illusion is there, but it still sounds off. Playing behind the beat is a feeling; it's not a calculation. As a drummer keeps time, the band reacts by lo…
That feeling of connection with other musicians while playing is something else.
I was live-reading lyrics sheets to some songs I’d never heard while jamming with a big group. Hit a chorus with some really great phrasing and bungled it the first time through. But the second time, the other guy singing and I just automatically made eye contact and had a whole conversation through body language.
“I’ve got it this time”
“Yeah?”
“Yeah”
“Oh fuck yeah”
“Fuck yeah indeed, my good fellow”
I've played guitar for 23 years, and there is something just off-putting about most of the music on that page, but particularly "Yellow Bus Jam". The guitar solo sounds very unnatural, especially the phrasing, which is totally random. Blues musicians are actually attempting to say something through their instrument. This was just a random number generated solo played by a 6 finger three handed robot. No thanks, lol.
Does it sound like something a human could play? You're not attacking how it sounds but what it's playing.
I've played guitar for 23 years, and there is something just off-putting about most of the music on that page, but particularly "Yellow Bus Jam". The guitar solo sounds very unnatural, especially the phrasing, which is totally random. Blues musicians are actually attempting to say something through their instrument. This was just a random number generated solo played by a 6 finger three handed robot. No thanks, lol.
Still makes me want to give up. I just started learning keyboard and playing with synthesizers in the last 6 months or so with the intention of making game music and it's tough to not feel like I'm wasting my time. Game devs will go with what they can afford and who can blame them? The output is not perfect, but if the GenAI can do this now, what will it sound like a year from now? Two? Really takes the wind out of the sails of us newbies.
This music sounds like the visual equivalent of one of those motion settings on a tv that cause the "Soap Opera Effect". I'm wondering if the LLM is failing to emulate the odd and even harmonics of real recordings and that's why it seems so unnatural. It makes me feel sick. It's different than the sterile straight to digital recordings of the late 80s early 90s too. Do any of these let people download the individual…
> I'm wondering if the LLM is failing to emulate the odd and even harmonics of real recordings What do you mean? If you screw up even/odd harmonics it will sound far less natural, like confusing flutes and clarinets.
I meant as an example if someone is singing, holding a long note and their voice is producing a lot of even harmonic overtones but every few cycles that overtone changes randomly flips or disappears. Sonically disoriented.